[sword-devel] linking syntax

Chris Little chrislit at crosswire.org
Sat Nov 29 02:26:35 MST 2008


Troy A. Griffitts wrote:
> The pros in the schema are wrong.  I'm sure I've asked Patrick to change 
> them.  The actual spec has:
> 
> <xs:attribute name="osisRefWork" type="osisWorkType" use="optional" 
> default="Bible"/>

Ok. Prose errors aren't unheard of.

A second issue you raised, concerning standard meta workID values: What 
should we go with?

You mention:
bible:
strong:
self:


For the first, I'm going to argue for Bible:. The spec uses it, as you 
quoted. I've used it in a lot of shipping content. And, inevitably 
someone will come along and ask why we have Bible lowercased--we'll say 
"it's our camel-casing standard" and they'll respond "so hate Jesus?"

For the second, I have no preference, so long as we pick & standardize 
on one. I think the filters recognize half a dozen options. There's a 
"Strong" module now/soon, which might be an argument against "Strong:", 
since then "Strong:" could point to that module specifically, while 
"strong:" (for example) points to the user's preferred lexicon for 
Strong's numbers (Strong, NASlex, etc.). The manual includes an instance 
of "s:" but I suspect Todd is the source of that. That's concise, if not 
necessarily clear. And "strong:" of course is very clear.

For the last, I'm sure I've seen "this:". But again that might be Todd's 
work rather than anything we need necessarily follow. Programmers will 
definitely understand it. Non-programmers might better understand 
"self:" as you propose.

(To summarize my opinions: I strongly favor "Bible:", weakly favor 
"strong:" (strongly disfavor "Strong:", and have no opinion on "self:" 
vs. "this:".)

Other opinions, from encoders especially, and particularly those who may 
have used any of the above, would be welcomed.

--Chris




More information about the sword-devel mailing list